Whenever we think of Data Centre scale, many of us think of huge warehouses with rows and rows full of racks of servers – and in some cases we are right. They can range from a few servers in a room to large structures measuring in hundreds of thousands of square feet with tens of thousands of servers and other hardware. The sizes and types of hardware they contain vary depending upon the needs of the entity or entities they are supporting.
There are many different types such as private cloud providers like the colo's, public cloud providers like Google or Dropbox, organisations’ private data centres and Government Data Centres such as the UK Government or various scientific research facilities. These various Data Centres are not staffed like your work office with one person per computer, but with a smaller number of people monitoring large numbers of computers and networking devices, as well as power, cooling and other important building facilities. The floors must withstand more weight than a typical office building because Data Centre equipment is by no means light. On top of this, they also require high ceilings to accommodate equipment like talk racks, raised floors and ceiling-hung cabling, among other things.
Many companies with large online presence have large Data Centres located all over the world, this could include companies such as Google, Facebook and Microsoft. Microsoft for example reportedly adds 20,000 servers monthly as of 2017. Google has more than a dozen large Data Centres in various countries such as Ireland, America, Hong Kong and Finland.
(Google's Data Centre Located In The Dalles, Oregon)
The configuration of servers, network systems and the supporting equipment can vary greatly depending upon the organisation, purpose, location, growth rate and the initial design of the Data Centre. The layout can greatly affect the efficiency of data flow and the environmental conditions within the Data Centre. Some sites might divide their servers into groups by functions, such as dividing web servers, application servers and database servers.
There are no hard and fast rules and there aren’t many official standards. However, some groups are trying to create guidelines. The Telecommunication Industry Association (TIA) developed a Data Centre tier classification standard in 2005 called the TIA-942 project which identified the four categories of the Data Centre. Categorised by metrics like redundancy and lever of fault tolerance. The tiers are the following;
- Tier 1 – This tier is a basic site infrastructure with one distribution path that has no built-in redundancy.
- Tier 2 – This tier is a redundant site infrastructure with a single distribution path that includes redundant components.
- Tier 3 – This tier is a Simultaneously maintainable site that has multiple paths (only one of which is active at a chosen time).
- Tier 4 – This tier is a fault tolerant site that has multiple active distribution paths for a range of redundancy.
Typically, sites that fall into both Tier 1 and Tier 2 have to shutdown for a maintenance occasionally where as Tier 3 and Tier 4 sites should be able to stay up during maintenance and other interruptions. The standard also spells out recommendations for various factors such as cabling, facilities, infrastructure, air conditioning. These are aimed for the telecommunications industry but can be applied to other Data Centres. It is one of the few ways we can rate and compare Data Centres by overall design and functionality.